Optimisasi Model Regresi Linier Menggunakan Pendekatan Teori Rough Set
DOI:
https://doi.org/10.30983/lattice.v5i2.10376Keywords:
Rough Set Theory, Regresi Linier Berganda, Data ReductionAbstract
Linear regression is widely used to model student performance data; however, its effectiveness can decrease when applied to datasets containing inconsistent samples, which affects the clarity and stability of the model. This study explores the use of Rough Set Theory (RST) as a data reduction approach to improve the quality of linear regression modeling. RST is applied in the pre-modeling stage to identify and reduce inconsistent samples through two schemes: majority-keep reduction and strict reduction. Linear regression models are then built using the reduced datasets and compared with the initial model based on the coefficient of determination (R²) and classical regression assumption tests. The results show an increase in R² from 0.624 in the initial model to 0.741 with RST majority-keep and to 0.862 with RST strict reduction, indicating improved model fit after data reduction, and the classical regression assumptions are satisfied. These findings suggest that integrating RST improves the diagnostic quality and stability of linear regression, with majority-keep reduction providing an optimal balance between enhancing model and maintaining a representative sample size.
Regresi linier banyak digunakan untuk memodelkan data student performance. Namun, efektivitasnya dapat menurun ketika diterapkan pada data yang mengandung sampel inkonsisten sehingga memengaruhi kejelasan dan kestabilan model. Penelitian ini mengkaji penggunaan Rough Set Theory (RST) sebagai pendekatan reduksi data untuk meningkatkan kualitas pemodelan regresi linier pada data student performance. RST diterapkan pada tahap pra-pemodelan untuk mengidentifikasi dan mereduksi sampel yang tidak konsisten melalui dua skema reduksi, yaitu majority-keep reduction dan strict reduction. Model regresi linier kemudian dibangun menggunakan dataset hasil reduksi dan dibandingkan dengan model awal berdasarkan nilai koefisien determinasi (R²) dan hasil pengujian asumsi klasik regresi. Hasil penelitian menunjukkan bahwa nilai R² meningkat dari 0,624 pada model awal menjadi 0,741 pada model dengan RST majority-keep dan 0,862 pada model dengan RST strict reduction, yang menunjukkan peningkatan kecocokan model pada data yang dianalisis setelah dilakukan reduksi data dan uji asumsi klasik terpenuhi. Analisis ini mengindikasikan bahwa integrasi RST berkontribusi pada peningkatan kualitas diagnostik dan stabilitas model regresi linier melalui reduksi data. Di antara kedua skema reduksi, RST majority-keep memberikan keseimbangan yang lebih baik antara perbaikan model dan mempertahankan ukuran sampel yang representatif.
References
D. H. Maulud and A. M. Abdulazeez, “A Review on Linear Regression Comprehensive in Machine Learning,” vol. 01, no. 02, pp. 140–147, 2020, doi: 10.38094/jastt1457.
K. Qu, “Research on linear regression algorithm,” vol. 01046, 2024.
M. A. Iqbal, “Aplication of Regression Techniques with their Advantages and Disadvantages.” 2021.
N. Roustaei, “Application and interpretation of linear-regression analysis,” 2024.
Y. Abdulraheem, “Journal of Public Health Issues and Practices The Role of Regression Analysis in Preventive Research Modalities : A Medical-Focused Comprehensive Review,” vol. 9, pp. 1–5, 2025.
S. Chung, Y. W. Park, and T. Cheong, “A mathematical programming approach for integrated multiple linear regression subset selection and validation,” Pattern Recognit., vol. 108, 2020, doi: 10.1016/j.patcog.2020.107565.
Rasyidah, R. Efendi, N. M. Nawi, M. M. Deris, and S. M. A. Burney, “Cleansing of inconsistent sample in linear regression model based on rough sets theory,” Syst. Soft Comput., vol. 5, no. December 2022, 2023, doi: 10.1016/j.sasc.2022.200046.
Z. Pawlak, “RoughSets,” Inst. Theor. Appl. Informatics, Polish Acad. Sci. ul. Bałtycka 5, 44 100 Gliwice, Pol., vol. 2, pp. 1–51, 1982.
S.Santoso, Statistik non-parametrik. Elex Media Kompotindo, 2001.
R. Słowiński, J. Stefanowski, S. Greco, and B. Matarazzo, “Rough set based processing of inconsistent information in decision analysis,” Control and Cybernetics, vol. 29, no. 1. pp. 378–404, 2000.
D. Dubois and H. Prade, “Rough fuzzy sets and fuzzy rough sets,” Int. J. Gen. Syst., vol. 17, no. 2–3, pp. 191–209, 1990, doi: 10.1080/03081079008935107.
J. Komorowski, L. Polkowski, and A. Skowron, “Rough sets: A tutorial,” Rough fuzzy Hybrid. A new trend Decis., pp. 3–98, 1999, [Online]. Available: http://secs.ceas.uc.edu/~mazlack/dbm.w2011/Komorowski.RoughSets.tutor.pdf
I. Rahmi, R. Efendi, and N. A. Samat, “Examining Risk Factors of Anemia in Pregnancy Using,” BAREKENG J. Math. App, vol. 18, no. 1, pp. 537–552, 2024.
X. Su, X. Yan, and C. L. Tsai, “Linear regression,” Wiley Interdiscip. Rev. Comput. Stat., vol. 4, no. 3, pp. 275–294, 2012, doi: 10.1002/wics.1198.
I. Rahmi, R. Efendi, N. A. Samat, H. Yozza, and M. Wahyudi, “The Effects of Data Reduction Using Rough Set Theory on Logistic Regression Model,” Lecture Notes in Networks and Systems, vol. 1078 LNNS. pp. 64–73, 2024. doi: 10.1007/978-3-031-66965-1_7.
H. Kim and H. Kim, “Statistical notes for clinical researchers : simple linear regression 3 – residual analysis,” vol. 44, no. 1, pp. 1–8, 2019.
N. Shrestha, “Detecting Multicollinearity in Regression Analysis,” no. June, pp. 1–5, 2020, doi: 10.12691/ajams-8-2-1.
G. Vining, E. A. Peck, and D. Montgomery, Introduction to Linear Regression Analysis. 2012.
R. Efendi, S. Mu’at, V. A. Dewi, N. Arisandy, N. A. Samsudin, and D. S. S. Sahid, “Rough-regression for categorical data prediction based on case study,” Proceeding - 2019 Int. Conf. Artif. Intell. Inf. Technol. ICAIIT 2019, pp. 277–281, 2019, doi: 10.1109/ICAIIT.2019.8834584.
Downloads
Published
How to Cite
Issue
Section
Citation Check
License
Copyright (c) 2025 Lita Lovia, Yessy Yusnita, Izzati Rahmi, Widdya Rahmalina

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with Lattice Journal : Journal of Mathematics Education and Applied agree to the following terms: Authors retain copyright and grant the Lattice Journal : Journal of Mathematics Education and Applied right of first publication with the work simultaneously licensed under a Creative Commons Attribution License (CC BY-SA 4.0) that allows others to share (copy and redistribute the material in any medium or format) and adapt (remix, transform, and build upon the material) the work for any purpose, even commercially with an acknowledgement of the work's authorship and initial publication in Lattice Journal : Journal of Mathematics Education and Applied. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in Lattice Journal : Journal of Mathematics Education and Applied. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
